Aspects of Continuous Deployment – A Multi-Dimensional View
summary
Swiss Post’s parcel sorting decision system runs in AWS Frankfurt on AKS with Kubernetes, ~20 services (mostly containers), .NET, Kafka (≈70 topics) and MQTT, processing ~300M parcels/year (peak >2M/day) and 400–500 GB of state; team ships ~3,500 deployments/year (~10/day), records 4 change failures with ~1 hour restore, and enforces 100% baseline unit tests and a zero warnings policy. They cut pipelines toward 6‑minute CI runs, mentor juniors to raise technical excellence, and claim the system works in production.
AI-written summary · about
chapters (12)
- 0:00Continuous deployment perspectives
- 3:08Three to 3,500 deployments/year
- 8:11Swiss Post parcel sorting
- 13:31Team excitement about system
- 16:43Concrete vision
- 20:17Local purpose to goals
- 25:41Quality and speed in development
- 30:14Technical excellence in teams
- 35:17Mentoring and learning
- 38:08Ownership in continuous deployment
- 40:38Technology vs team dynamics
- 41:48Q&A
related talks
48mA Journey to Safe, Mission Critical Continuous Delivery at Descartes Labs
54mFrom Code to Customer
41mAll Things Continuous
40mBalancing agility and security: SBB’s journey to AWS(SEC204)
40mContinuously Deliver your Kubernetes Infrastructure
36mThe journey towards K8s at Deutsche Bahn
51mContinuous Deployment for a Billion Dollar Order System
23mBackbone of the Swiss Financial Center Goes DevOps